This prospective study will determine if uncontrolled seizures over time impair development by comparing measures of morphometry, cognition, thought disorder, and psychopathology in 6 - 18 year old children with complex partial seizure disorder (CPS), primary generalized epilepsy with absences (PGE), and normal children at baseline and 2 years later. By including nonepileptic siblings of the CPS, PGE, and normal subjects, the project will ascertain if the hypothesized developmental abnormalities in the patients are epilepsy-related rather than familial. Specific Aims: The study will test 3 main hypotheses: (a) CPS and PGE patients with an increase in seizure frequency from baseline through the 2 year follow-up will have a drop in IQ and an associated increase in thought disorder and psychopathology compared to patients with a decrease in seizure frequency and the normal subjects. (b) The patients with an increase in seizure frequency will have a smaller age-related increase in mesial temporal lobe (i.e., hippocampus, amygdala) and frontal lobe white matter volumes than the patients with a decrease in seizure frequency and the normal subjects. (c) The smaller age-related increase in mesial temporal lobe and frontal lobe white matter volumes will be associated with the predicted drop in IQ, and increase in thought disorder and psychopathology in the patients with uncontrolled seizures. Significance: By addressing the on-going debate "Do seizures impair development?" the study's findings will delineate neurobiologic mechanisms of impaired cognition, thought disorder, and psychopathology in middle childhood epilepsy. They will underscore the importance of medical control of seizures to promote optimal development of cognition, communication, and behavior in these children. They will highlight the potential use of MRI for identifying children at risk for thought disorder and pave the way for future studies on treatment and prevention of thought disorder and psychopathology in pediatric CPS and PGE. Research Design and Methods: The study mill recruit 40 CPS, 40 PGE, and 40 normal children matched on age, gender, ethnicity, and socioeconomic status. Twenty-five siblings of the CPS, PGE, and normal subjects will also participate in the study. Each subject will undergo an MRI, as well as cognitive, thought disorder, behavioral, and linguistic testing at baseline and 2 years later. Weekly seizure frequency data will be obtained on a monthly basis from the parent during the 2 years.